Getting Away With Murder: Homicide Clearance by Arrest in Chicago and Its Community Areas

Abstract

This work explores the decline in murder clearances through arrest in Chicago from 1965 through 2015, specifically focusing on the most recent time period since 2001. The findings suggest that clearance by arrest has decreased significantly, that elapsed time is a limited factor in clearing more murders through arrest and that factors associated with clearance by arrest in Chicago have changed over time. These results lead to a discussion on the missing variance that cannot explain murder clearance by arrest as well as future research areas that can explore why many murderers in Chicago are increasingly escaping the justice system.

Keywords

Chicago homicide clearance policing arrest homicide clearance factors

Introduction

Murder¹ is a part of life in Chicago, and has been for a long time. In the 61 years between 1957 and 2017, 38,950 people were murdered in Chicago, an average of 639 per year ranging from 296 in 1957 to 970 in 1974². The average annual murder rate in Chicago between 1985 and 2014 was 22.5 compared to 14.7 for all other cities with at least 1 million people over the same period (FBI, n.d.). In 2016, Chicago experienced 765 murders, 17.9% higher than New York City and Los Angeles combined (628 total murders), despite having 360% fewer residents. Nationally, murder clearance percentages have declined from around 90% in the 1960’s to about 60% at the turn of the century (Litwin, 2004; Riedel, 2008; Riedel & Jarvis, 1999). The same trend is seen in Chicago as the murder clearance percentage through arrest³ was 91.7% (93.9% overall clearance) in 1965 and 57.7% in 1995 (Block et al., 2005); by 2005 just 195 of the 448 murders were cleared (43.5%). Thus, murder clearance has declined precipitously over time, limiting the deterrent effects of arrest, leaving many victims without justice, and rendering society less safe because those capable of murder remain free in communities across the country (Mancik et al., 2018; Ousey & Lee, 2010; Riedel & Boulahanis, 2007; Riedel & Jarvis, 1999; Riedel & Rinehart, 1996).

Surprisingly, in-depth empirical research into homicide clearance in general is only about two decades old despite the attention homicide generally receives (starting with, as examples: Borg & Parker, 2001; Puckett & Lundman, 2003; Regoeczi et al., 2000). Most of this early research focused on the characteristics of homicide clearance rather than exploring its decline (Xu, 2008). The current study explores the changes in murder clearance through arrest in Chicago and its 77 officially defined Community Areas since the turn of the century in light of its relatively high volume of murders. This research focuses on three questions: (1) whether homicide clearance by arrest in Chicago has declined significantly over time, and if so, to what degree; (2) what impact does time have on the ability to clear older cases through arrest; and (3) what factors, at the Community Area level, significantly explain homicide clearance by arrest between 2001 and 2015?

What Is Known About Homicide Clearance in Chicago Over Time

Perhaps no other city, sadly, has had as much homicide research conducted on it as Chicago. A more limited yet specific body of literature on homicide clearance in the Windy City also exists. This research largely derives from the existence and availability of a database containing information on all Chicago homicides that occurred between 1965 and 1995 created by Block et al. (2005), though other databases have also been utililized in a few studies. This literature is summarized below.

Riedel and Rinehart (1996) studied Chicago clearance rates for all homicides between 1987 and 1991. In this study, a cleared case was one with complete information included for the offender’s age, race and gender, that is, an identified offender. The main findings were that victim gender, age, race, and weapon type used on their own were not significant predictors of a homicide being cleared in Chicago over the 5-year period. By contrast, when a murder was part of another felony, it was uncleared 58% of the time, compared to 10% of murders that were not part of another felony. In this work, a felony circumstance for the murder was defined as the murder occurring during commission of another felony (e.g., robbery or rape) while a non-felony circumstance was defined as a murder being the crime committed (i.e., arguments, drunken brawls). The authors concluded that victim characteristics did not affect the probability that a homicide would be cleared in Chicago, an outcome that was related more to circumstances and the interaction of other variables, such as change in lifestyle as people age. This early evidence in Chicago pointed away from police devaluation of victims and toward event factors at the individual level of analysis.

Litwin (2004) utilized arrest as the clearance measure for his study of Chicago homicide outcomes from 1989 through 1991, similar to the time frame utilized by Riedel and Rinehart (1996). This study tested two competing theories of homicide clearance: Black’s theory of uneven law application and Klinger’s theory that police value all homicide victims equally. Using 2,224 cases across 72 Chicago Community Areas from a multilevel perspective, Litwin found that clearance rates were not different for black victims compared to whites, but homicides with Hispanic victims were significantly less likely than white victims to be cleared. The latter disparity could be explained by other factors beyond police discretion. Furthermore, homicides involving older victims were less likely to be cleared than those with younger victims. Homicides of victims with prior arrest records were as likely to be cleared as those with victims who had no criminal record. These results led Litwin to conclude little support for Black’s theory of the uneven application of law in homicide clearance in Chicago, though the data used is now 30 years old and these factors change over time (Jiao, 2007).

Litwin also studied non-discretionary factors which impacted homicide clearance. Victims found inside residences were significantly more likely to be cleared than those found in general public areas. Victims shot were less likely to lead to clearance than other weapons while cases involving other felonies (e.g., during a robbery) were less likely to be cleared than other circumstances. Homicides that occurred in 1991 were less likely to be cleared than those that occurred in 1989, showing the homicide clearance rate could be a result of internal changes within the CPD. At the wider community level, a community area’s median income was not a factor in clearance, nor was unemployment or educational levels, whereas home ownership rates were a significant and positive factor in homicide clearance. Finally, neither an area’s population nor homicide rate were significant predictors of homicide clearance. These results led Litwin (2004, p. 345) to conclude that “overwhelming support exists for the importance of nondiscretionary factors in understanding homicide clearances in a given time period,” more reflective of the changing nature of homicide itself impacting clearance rates in Chicago more than discretionary or community level factors. As a result, Litwin suggested replicating this study over time in Chicago, what the present study seeks to do.

Litwin and Xu (2007) studied whether changes over time were related to homicide clearance in Chicago. The authors used three distinct time periods—1966 to 1975, 1976 to 1985, and 1986 to 1995—to determine which factors were most associated with a decline in clearance rates using multilevel modeling. They found that clearance declined from 88.6%, to 80.7% and then 72.5% over the time span. At the victim level, victim age was a significant predictor of clearance (the younger the victim, the higher the clearance rate), as was the victim being black or Hispanic (lower clearance rates than whites). The victim’s prior arrest record was only significant in the earliest time period, but body location (for those found inside home locations) was consistent across all three time periods. Community Area characteristics were also found to be significant and their effect was not constant, varying over time. Economic disadvantage became significant in the last time period as did the percentage of Spanish-speaking population, but black population was not significant in any time period. Other community area variables were not significant in any period. These authors concluded that Community Area characteristics are a factor but to a lesser degree than victim and situational characteristics, consistent with Litwin’s (2004) research.

Using the same data over the same time period (1965–1995), Jiao (2007) calculated the overall homicide clearance rate at 81%. Victim race was not a significant predictor of clearance once control variables were introduced into the logistic regression models, and the clearance rate significantly declined over time as a function of year measured continuously. Cases that occurred in indoor locations other than the home were cleared significantly more likely to be cleared than those in street and outdoor locations, but cases that occurred in vehicles or public transportation systems were cleared significantly less than incidents that took place in the street or outdoors. Homicides involving known relationships as family members, friends, and gang/business associates were cleared significantly more than those involving strangers or no known relationships. This research did not include Community Area factors in the analysis.

The decline in homicide clearance in Chicago was also studied from a prosecutorial perspective in Chicago from 1988 through 1995. Riedel and Boulahanis (2007) studied homicides cleared by exceptional means rather than arrest or the combination of the two measures, specifically focusing on those cases which “encountered a circumstance outside the control of law enforcement that prohibits the agency from arresting, charging, and prosecuting the offender” (p. 153), which is the same definition used by the CPD (Chicago Police Department, 2011). The authors found that the total clearance rate in Chicago between 1982 and 1995 was 75.3%, 10.7% higher than the 64.6% cleared by arrest, suggesting that there is value in distinguishing between the two measures that are often combined into one clearance percentage. This study found that certain types of cases were more prevelant for being barred for prosecution than others, meaning those decisions also impact the overall clearance rate.

Xu (2008) used the same data to study homicide events and the decline in homicide clearance (measured as both arrest and exceptional) in Chicago between 1966 and 1995, using a longitudinal approach for the latter at the individual and community area level of analyses. This study found that in Chicago, “easier” to clear homicide cases (younger victims and homicides within homes) actually increased in proportion between 1965 and 1995 where concomitant felony cases declined proportionally after peaking in 1974. However, even clearance rates for easier to solve homicides declined over the 30-year period: clearance rates for homicides committed inside homes decreased from 93% in 1967 to 50% in 1995. Exceptional clearances amounted to less than 10% of all cases cleared, consistent with previous research (Riedel & Boulahanis, 2007). Similar to Jiao (2007), Xu found that factors contributing to the decline in homicide clearance in Chicago changed over time. Several victim characteristics aggregated to the community level—age, sex and prior record—were not significant predictors of the decline in homicide clearance. While the percentage of black victims was significant in three of the four models, the percentage of Hispanic victims was significant in all four models. The increase in the proportion of stranger homicides also contributed significantly to the decline, as did the percentage of bodies found in vehicles and an increase in the proportion of Hispanic victims. Furthermore, the clearance rate for both firearm and non-firearm homicides both declined, suggesting that changes in the use of firearm as the weapon of choice in homicides may not be a major factor in the clearance decline. Community level variables such as home ownership and vacant housing were not significant factors of the Chicago homicide clearance decline.

Alderden and Lavery (2007) analyzed homicide clearance in Chicago from 1991 to 2002 using an internal CPD dataset, the first departure from the dataset used in the previous research. This new data permitted analysis using different variables in a more recent time period. The authors found that 48.5% of cases cleared were done so within 7 days of the homicide while 95.7% of cases cleared were cleared within 2 years, showing time to be a factor. Further analysis revealed that roughly 20% of homicides had causes unknown to detectives, and when these unknowns were included in the sample, the clearance rate was 66.3% compared to 78.8% when they were eliminated. Using logistic regression, this study found that individual characteristics did matter in Chicago relative to homicide clearance. Cases involving female victims were significantly more likely to be cleared than those involving male victims (in the full model), as were cases involving victims <10 years old or 10 to 25 years old compared to those older than 26. Consistent with previous research, it was also found that while there was no difference in clearance for white and black victims, cases involving Hispanic victims were significantly less likely than white victims to be cleared. Cases involving firearms were significantly less likely to be cleared, as were cases where the victim was found outdoors compared to indoors. Specifically, when the homicide was gang-related, victim characteristics were significant predictors of clearance: black and Hispanic victims less likely than white victims to lead to clearance while cases where the victim had no prior arrest were more likely to be cleared than victims with prior arrests (Alderden & Lavery, 2007).

Mancik et al. (2018) also utilized different data sources to study homicide clearance in Chicago between 1996 and 2000 as as function of collective efficacy at the neighborhood cluster level (between census tract and community area geographies). Using data from the CPD, the Census Bureau and the Project on Human Development in Chicago Neighborhoods, the study assessed overall homicide clearance in 321 neighborhood clusters, finding a clearance rate of 67.7%. Due to issues of multicollinearity between the census variables, the authors created three indexes: economic disadvantage, immigrant concentration, and residential stability. Their initial negative binomial model found that neighborhood economic disadvantage and residential stability were related to lower clearance rates, while immigrant concentration, legal cynicism and specific case characteristics (percent white victim, percent male victim, percent stranger, percent firearm, and percent in residence) were not significant predictors of clearance at the neighborhood level. Collective efficacy was a significant predictor of homicide clearance, though neighborhood economic disadvantage and residential stability remained significant. Neighborhood victimization, as measured by a survey question that asked about prior 6-month victimization, was a significant yet weak factor in all three models. The final model excluded low homicide neighborhood clusters, changing the findings somewhat. A few case specific characteristics (greater percentage of white victims and greater percentages of firearm related homicides) corresponded with increases in the clearance rate. The results led the authors to conclude that neighborhood context did matter in Chicago relative to homicide clearance.

Finally, Ferrandino (2018) studied Chicago homicides epidemiologically, focusing on the homicide spike in 2016, but also touching on homicide clearance by arrest at the block group level in Chicago from 2001 to 2015. His work found that clearance by arrest over this time period was highest (53.4%) in the block groups most afflicted by homicide. These 59 block groups (3% of city population with 14.8% of the city’s homicides) were 90.4% Black and 6.1% Hispanic. The lowest homicide clearance by arrest percentages (38.7%) were found in the 1,148 block groups where homicides were present but non-adjacent to afflicted block groups (56% of the city’s population). These block groups were far more diverse racially: 24.8% White, 32.6% Black, 5% Asian, and 36% Hispanic. Due to the focus of that research, these findings were merely descriptive and not included in further modeling relative to homicide clearance by arrest.

Although the results in Chicago were mixed depending on the data source, the unit of analysis and the perspective taken, the overarching findings were that homicide clearance is lower in more recent periods compared with the past. This is consistent with previous research generally (Riedel, 2008; Riedel & Jarvis, 1999), though only two Chicago studies focused solely on clearance through arrest (Ferrandino, 2018; Litwin, 2004). The present study uses several sources of data over time to assess homicide clearance through arrest in Chicago.

Most of the research reviewed from Chicago shows little support for the police devaluation perspective, often framed through Black’s perspective of law (Ferrandino, 2018; Jiao, 2007; Litwin, 2004; Litwin & Xu, 2007; Riedel & Rinehart, 1996; Xu, 2008) but this was more nuanced when internal CPD data was used and the focus was on gang-involved shootings (Alderden & Lavery, 2007). The caveat is that the Chicago research has specifically found that Hispanic victims and populations are associated with lower clearance rates, a finding that is consistent with other samples at the national or large city level of analysis (see Peterson, 2015; Riedel, 2008; Roberts & Lyons, 2011 as examples). This finding does not mean that CPD devalues Hispanic victims, just that it is a significant factor relative to lower homicide clearance in the city.

The event characteristics perspective is defined as “variation in homicide clearance is a function of homicide incident characteristics that complicate the investigation, as opposed to any factors associated with selective efforts on behalf of law enforcement” (Rydberg & Pizarro, 2014, p. 343). This perspective did find support in Chicago relative to homicide clearance rates (Alderden & Lavery, 2007; Jiao, 2007; Mancik et al., 2018; Riedel & Rinehart, 1996; Xu, 2008). These findings are consistent with broader homicide clearance research that details the changing nature of homicide impacts investigative outcomes, especially related to firearm-involved homicides, which profligate in Chicago (Baskin & Sommers, 2010; Ousey & Lee, 2010). The results for community level factors in Chicago were also mixed, but generally weakly related to homicide clearance (Litwin, 2004; Litwin & Xu, 2007; Mancik et al., 2018; Xu, 2008). The Chicago findings are more muted than previous research generally which finds significant variation in homicide clearance rates within individual cities at the neighborhood level (Peterson, 2015; Regoeczi & Jarvis, 2013) as well as across cities at the national level (Borg & Parker, 2001).

One major gap in the previous research that the present study seeks to fill is the use of more recent data (2001–2015) to draw conclusions about homicide clearance specific to one large, violent city. Previous research into homicide clearance in Chicago has largely utilized the Homicide in Chicago data 1965 to 1995 created by Block et al. (2005), Jiao (2007), Litwin (2004), Litwin and Yu (2007), Riedel and Boulahanis (2007), Riedel and Rinehart (1996), and Xu (2008). Other research has used internal Chicago police data either in isolation or combined with other data sources, but neither of these studies extended the data past 2002 (Alderden & Lavery, 2007; Mancik et al., 2018). Though Ferrandino (2018) did use the most recent time period in Chicago, his focus on this topic was largely descriptive. As Chicago has been experiencing increasing incidents of homicides and shootings over the past several years, more recent data is needed to determine if clearance factors have changed and what factors remain significant.

Data and Methods

This section details the data and methods used to answer the three research questions guiding this analysis: (1) whether homicide clearance by arrest in Chicago has declined significantly over time, and if so, to what degree; (2) what impact does time have on the ability to clear older cases through arrest; and, (3) what factors, at the Community Area level, significantly explain homicide clearance by arrest between 2001 and 2015?

This study utilizes the Community Areas in Chicago as the unit of analysis. Since the 1920’s, Chicago has been officially delineated into 77 Community Areas, whose boundaries have remained the same since (Encyclopedia of Chicago, n.d.). The spaces are geographical, not tied to populations or specific groups nor do they represent specific cultural neighborhoods. The benefit of this approach is that data is consistently collected in these areas over time. This unit of analysis has also been used in previous homicide research in Chicago.

The data for this research originates from several sources. The first is the Chicago Data Portal Crimes—2001 to Present Database (Chicago Data Portal, 2020). This dataset is publicly available and contains a wealth of data on every crime reported to the Chicago Police Department since January 1, 2001 and extracted from the CLEAR (the Citizen Law Enforcement Analysis and Reporting) record management system. The only crime delineated by victim are homicides (entry for each victim rather than each incident). As a result, the focus is on total homicide victims, not homicide events. This is the first known study to use the public open data source to research homicide clearance in Chicago.

For this research, the description column was filtered for first degree murder only, creating a database of all first degree murders in Chicago since 2001, as defined by the CPD using the Illinois Uniform Crime Reports (I-UCR, 2017) definitions. The I-UCR (2017, p. 2) defines criminal homicide as “the willful killing of one human being by another or the killing of another person through gross negligence. As a general rule, any death caused by injuries received in a fight, argument, quarrel, battery, or commission of a crime must be reported as a criminal homicide.” This can be coded as in the first or second degree upon occurrence. The variables in the dataset include: the case number, date the murder was reported, truncated address where it occurred (last two numbers of the street address removed 1XX Main Street, as an example), UCR code, offense type (homicide, robbery), offense description (first-degree murder, armed robbery), location type (i.e., alley, street, CTA station), whether an arrest was made (yes, no), whether the event was related to a domestic situation (yes, no), beat, district, ward, neighborhood, XY coordinates of the murder location, year murdered occurred, and last date the record was updated. The filtered data (for only crimes classified as first-degree murder) was downloaded on three separate occasions to assess the change of arrests over time, making the data fluid. The original dataset was downloaded on April 1, 2016; the second dataset was downloaded on August 27, 2018 (about 28 months later); and the final dataset was downloaded March 5, 2019 (about 35 months after first dataset and 6 months after the second update). This process permitted analysis that partially overcomes one of the main methodological concerns about using annual arrest and murder totals to calculate clearance; namely, taking total arrests and dividing by the total number of homicides regardless of when the arrest occurred. As the dataset utilized for the present study does not include clearance by exceptional means, only clearance through arrest is used to determine the effect of time on murder clearance. This data was disaggregated from the city to the community area levels.

To truly study the different measures of clearance over time, two other sources of data are used to supplement the Chicago Data Portal. The first is the Chicago offender data file for 1965 to 1995, which has been used in most of the prior studies on clearance rates and correlates in the city (Jiao, 2007; Litwin, 2004; Litwin & Yu, 2007; Riedel & Boulahanis, 2007; Riedel & Rinehart, 1996; Xu, 2008). The second data source is the Chicago Police Department 2011 Murder Analysis (Chicago Police Department, 2011), which provides data on all homicides from 1991 and 2010 that were cleared within each year. Thus, all three of these datasets combined give a holistic view on murder clearances in Chicago within year and over time to assess annual clearance trends from 1965 through 2015. The dependent variable is the first-degree murder arrest clearance percentage at the community level for each of the three most recent periods for 2001 through 2015.

The Chicago Data Portal (2017) also provides a critical source of data for several other study measures. The first is the Strategic Subject List, or SSL, a database utilized by the CPD to mathematically predict an arrestee’s probability of being a victim or offender in a shooting incident based on prior arrests made between August 1, 2012 and July 31, 2016. The dataset also contains several other variables that were combined to create a single summed index of gang, drug and gun violence offenders within each Community Area. This measure was created by summing the following variables into one index score for each Community Area: the percentage of arrestees with a gang affiliation, the average number of drug arrests per arrestee, the average number of arrests for unlawful use of weapons per arrestee, the average number of arrests for violent offenses per arrestee and the average number of times each arrestee has been shot or assaulted in the past. The inclusion of a gang index measure is consistent with previous research that finds gang-involved homicides are related to homicide clearance rates in Chicago (Alderden & Lavery, 2007) and is used as a Community Area level proxy for the gang, drug, and violence environment that cannot be directly measured and accounted for otherwise. The scores on this independent variable at the Community Area level ranged from 0.27 to 1.68, averaging 0.84.

Two language measures are used as independent variables, with both deriving from the Census Data—Languages Spoken in Chicago, 2008 to 2012 dataset (Chicago Data Portal, 2014a). The first is a measure of the total population of a given Community Area that is between 5 and 65 years old and speaks English less than well. The second measure is the percentage of this specific population that has Spanish as their primary language. These measures are consistent with research that has assessed the impact of immigrant concentration, measured as the percentage of foreign-born residents and those that speak Spanish (Mancik et al., 2018). Though immigration concentration was found not to be a significant predictor of homicide clearance in Chicago in that study, the authors note that few studies have included this variable and the data used in that study is now almost 25 years old. The inclusion of this variable helps to determine if their finding still holds true in Chicago. The second measure is consistent with research on Latino populations and victims relative to homicide clearance outcomes (Alderden & Lavery, 2007; Litwin, 2004; Litwon & Xu, 2007; Xu, 2008). The two measures, though seemingly similar, were not highly correlated (r = 0.29). The scores on the first measure averaged 15.6% but ranged from less than 1% to 44%; while the second measure averaged 58% and ranged from 0% to 99% at the Community Area level.

Socioeconomic status at the Community Area level is measured using the hardship index, calculated by the City of Chicago as an index of six census level population percentage variables—crowded housing, households below poverty, 16 years and older unemployed, ages 25 and older without high school degree, population under 16 and over 65 years of age, and per capita income—that range from the lowest (0) to highest (100) levels of hardship (Chicago Data Portal, 2014b). This measure is consistent with previous research (Litwin, 2004; Mancik et al., 2018). The hardship index averaged 49.5 and ranged from 1 to 99.

Race has also been studied as a factor affecting homicide clearance in Chicago at both the individual case and community levels (Alderden & Lavery, 2007; Jiao, 2007; Litwin, 2004; Litwin and Xu, 2007; Riedel & Rinehart, 1996; Xu, 2008). This study includes black and Hispanic population percentages at the Community Area level, calculated by the City of Chicago from Census data (Chicago Data Portal, 2020). Black population percentage averaged 39%, ranging from 0% to 98% while the Hispanic population percentage averaged 26%, ranging from 0% to 86%.

Two standardized measures of violent environment were created for this study using data from the initial Crimes 2001—Present Database (Chicago Data Portal, 2020). The first is a z-score of the average non-murder gun crime rate for 2008 to 2015 by Community Area. This is calculated by filtering any crime in the dataset that included a firearm or gun, excluding homicides, which totaled 81,157 for the city over the time period. The average is taken for the 8 years and converted to an average annual rate per 100,000 population and finally converted to a z-score. The second uses the same data but calculates the z-score for the total non-murder gun crime arrest percentage for 2007 to 2015 for each Community Area. This is a standardized measure of the general clearance rate for gun crimes that occurred in each Community Area between 2007 and 2015. Together these measures account for a Community Area’s broad non-homicide gun violence and clearance environment, similar conceptually to neighborhood victimization (Mancik et al., 2018).

The final measure, which is new to the study of homicide clearance in Chicago, has criminological roots related to the concept of fatalism. The variable is measured through the life expectancy in 2010 which derives from the Chicago Data Portal (2014c). This variable is useful in that Community Areas with lower life expectancies may respond to homicides differently as expressed in clearance through arrest, though this has not been included in the previous literature. Life expectancy across Community Areas averaged 77.6, ranging from 68.8 to 85.2.

To assess the first research question, patterns over time in arrest clearances for first-degree murder, a series of line graphs, three separate data sources are utilized in conjunction with descriptive statistics in tabular and line graph forms. This permits exploration of more recent clearance rates by arrest in Chicago than the previous body of literature. This includes three distinct time periods for the same data at intervals occurring after 2015 to determine how many cases are closed by arrest at further time frames. Furthermore, an analysis of variance was conducted comparing clearance rates by arrest at the community level over three distinct time periods (1965–1979, 1980–1995, and 2001–2015) to determine if clearance rates by arrest are significantly lower than, equal to or higher than the previous time periods.

To answer the second question, an analysis of variance was also conducted on the change in clearance rates over time for the all homicides that occurred between 2001 and 2015 at three time intervals afterward (April 2016, August 2018, and March 2019) to determine if that length of time made any significant impact on the ability to clear homicides in Chicago through arrest.

To answer the third question, ordinary least squares (OLS) regression is used at the Community Area level to determine what factors are significant in predicting the most recent time period (2001–2015) arrest clearance rate as of April 2016. These results are then compared with the previous literature to determine if the same findings hold over time or within Community Areas.

Results

Descriptive Analysis

Figures 1 to 3 below, from three different data sources over time, show the decline in murder clearance rates in Chicago graphically. Figure 1 uses data provided publicly by the CPD (2011) in its Annual Murder Report, and shows the percentage of homicides cleared (either exceptionally or by arrest) within each year from 1990 through 2011 in relation to total homicides. As seen in Figure 1, within year homicide clearance percentage by arrest decreased steadily over the time period, even as homicides declined concurrently. In 1990, about 65% of homicides were cleared within year compared to about 30% in 2011. Even as homicides decreased and technology utilization increased in Chicago, clearance rates steadily decreased.

The second source of data (Block et al., 2005) looks at overall by clearance type over time. This differs from within year data and does not suffer from a potential issue of lagged closure (homicide occurring in 1 year and cleared in another). Figure 2 shows a slight increase between 1965 (3.3%) and 1995 (7.4%) of exceptionally cleared cases, meaning this type of clearance was stable over time with a slight increase, though it did range from 2.9% to 14.6%. However, cases not cleared showed a steady increase over time, from 5% in 1965 to 34.9% in 1995 while cases cleared by arrest decreased steadily from 91.7% in 1965 to 57.7% in 1995. Thus, the decline in clearance rates is largely a result of a decline in the percentage of homicides that are cleared by arrest for homicides in Chicago.

The third source of data used is from the Chicago Data Portal for homicides that occurred between 2001 and 2015, updated for clearance changes at three intervals over time: April 2016, August 2018, and March 2019. As shown in Figure 3, the percentage of homicides cleared by arrest declined steadily over time. Interestingly, the graph also shows that each passing interval of time increases prior years’ clearance by arrest, suggesting that clearance rates are not fixed over time. The use of total arrests in a given year by total homicides that year is a poor measure (as discussed above), and it may be more beneficial to adjust prior year’s clearance totals to better reflect actual within year homicides that were cleared by arrest. Figure 3 also shows that this change extends back about 10 years. Previous research from Chicago using data from 1991 to 2002 found that 48.5% of cases cleared were done so within 7 days of the homicide, while 95.7% of cases cleared were cleared within 2 years. Regoeczi et al. (2008) studied the impact of time on homicide clearance, defining a cold case as one that took more than 2 weeks to solve. The updated data from Chicago show the trend toward fewer clearances by arrest, even several years after a homicide occurred. These findings suggest that time has not greatly improved homicide clearance through arrest.

Table 1 details the arrest percentages, at three points in time, for homicides which occurred from January 2001 through December 2015. As of April 2016, 51% of all homicides that occurred in Chicago since 2001 through the end of 2015 had been cleared by arrest, ranging from 63% of homicides in 2001 to 29% of homicides that occurred in 2015. This decline comports with the same findings from the broader literature and the other Chicago data in this study. In all, almost 40 months after the last homicide occurred (March, 2019), 139 additional homicides from 2001 through 2015 were cleared by arrest that were not cleared by December 2015, representing 1.9% of the 7,413 homicides in the sample. Of these 139 additional clearances, 60% were for homicides that occurred between 2011 and 2015. The trend continues, even more sharply, from 2016 through 2018.

Table 1.

Aggregate Homicides, Arrests and Percentage Cleared by Arrest, 2001 to 2018, Chicago, at Subsequent Time Intervals.*

Year	Homicides	Arrests by April 2016 (Time 1)	Arrests by August 2018 (Time 2)	Arrests by March 2019 (Time 3)	Arrest % by April 2016 (Time 1)	Arrest % by August 2018 (Time 2)	Arrest % by March 2019 (Time 3)
2001	668	423	423	425	0.63	0.63	0.64
2002	657	414	416	416	0.63	0.63	0.63
2003	602	367	368	369	0.61	0.61	0.61
2004	454	286	285	285	0.63	0.63	0.63
2005	452	269	269	269	0.60	0.60	0.60
2006	472	251	254	255	0.53	0.54	0.54
2007	449	242	246	247	0.54	0.55	0.55
2008	514	263	269	271	0.51	0.52	0.53
2009	461	217	224	226	0.47	0.49	0.49
2010	439	189	198	199	0.43	0.45	0.45
2011	437	177	181	184	0.41	0.41	0.42
2012	504	204	213	215	0.40	0.42	0.43
2013	419	168	183	183	0.40	0.44	0.44
2014	422	163	176	182	0.39	0.42	0.43
2015	463	133	177	179	0.29	0.38	0.39
Total 01–15	7,413	3,766	3,882	3,905	0.51	0.52	0.53
2016	778		236	245		0.30	0.32
2017	669		149	164		0.22	0.25
2018**	357		60	96		0.17	0.28
Total 16–18	1,804		445	505		0.25	0.28

Data derived from the Chicago Open Data Portal, Crimes 2001 – Present Dataset.

For homicides that occurred as of August 28, 2018, or Time 2.

One-Way Analysis of Variance

A one-way analysis of variance (not shown) was conducted to determine if any significant differences existed in the homicide arrest clearance rate at the community area unit of analysis (N = 77) over the three distinct time periods (1965–1979, 1980–1995, and 2001–2015). The data is shown in Table 2. The mean homicide arrest clearance rate overall (not within year) decreased from 0.83 (Period 1) to 0.65 (Period 2) and finally 0.57 (Period 3). The standard deviation decreased between the first two periods (0.13 to 0.09) but increased in the final time period (0.16), leading to equal variances not being assumed. The mean differences were significant (F = 84.06, p < .001, R² = 0.43, p = 1.0). As the variances were not equal (Levene’s < 0.05), the Tamhane’s statistic was used to interpret the post-hoc test, which found the homicide clearance rate to be significantly higher (p < .001) in the first time period compared to the second (MD = 0.19) and third time periods (MD = 0.26). The second time period also had a clearance rate that was significantly higher (p < .001) than the third time period (MD = 0.07). The results show that the visual trends seen in Figures 1 to 3 and displayed in Tables 1 and 2 are supported by the statistical analysis. Thus, arrest clearance percentages show a significant and substantial decrease over time in Chicago at the city and Community Area levels.

Table 2.

Homicides, Percentage Cleared by Arrest and Average Homicides Per Year, Chicago, by Community Area, 1965 to 2015.

Level	Homicides 1965–1979¹	Homicides 1980–1995¹	Homicides 2001–2015²	Homicide arrest % 65–79	Homicide arrest % 80–95	Homicide arrest % 01–15	Avg Hom/year 65–79	Avg Hom/year80–95	Avg Hom/year 01–15
Albany Park	27	100	53	1.00	0.64	0.64	1.80	6.25	3.53
Archer Heights	8	15	16	1.00	0.60	0.31	0.53	0.94	1.07
Armour Square	33	30	14	0.94	0.63	0.50	2.20	1.88	0.93
Ashburn	8	23	55	0.63	0.48	0.55	0.53	1.44	3.67
Auburn Gresham	200	332	241	0.84	0.61	0.43	13.33	20.75	16.07
Austin	375	780*	580*	0.85	0.63	0.51	25.00	48.75	38.67
Avalon Park	38	64	43	0.84	0.59	0.47	2.53	4.00	2.87
Avondale	32	74	64	0.84	0.70	0.58	2.13	4.63	4.27
Belmont Cragin	38	55	90	0.68	0.60	0.54	2.53	3.44	6.00
Beverly	8	25	7	0.75	0.68	0.57	0.53	1.56	0.47
Bridgeport	39	53	33	0.92	0.74	0.58	2.60	3.31	2.20
Brighton park	28	53	94	0.89	0.68	0.34	1.87	3.31	6.27
Burnside	12	16	25	0.75	0.81	0.60	0.80	1.00	1.67
Calumet Heights	28	59	42	0.82	0.59	0.45	1.87	3.69	2.80
Chatham	123	181	210	0.76	0.62	0.49	8.20	11.31	14.00
Chicago Lawn	16	127	218	0.88	0.70	0.45	1.07	7.94	14.53
Clearing	8	16	13	0.88	0.63	0.62	0.53	1.00	0.87
Douglas	234	314	65	0.97	0.64	0.62	15.60	19.63	4.33
Dunning	10	23	13	1.10	0.61	0.85	0.67	1.44	0.87
East Garfield Park	463	363	172	0.92	0.57	0.47	30.87	22.69	11.47
East Side	14	17	24	0.71	0.71	0.58	0.93	1.06	1.60
Edgewater	92	158	46	0.97	0.67	0.70	6.13	9.88	3.07
Edison Park	0	2	1			1.00	0.00	0.13	0.07
Englewood	602*	515	324*	0.89	0.66	0.44	40.13	32.19	21.60
Forest Glen	4	9	4	0.75	0.22	0.75	0.27	0.56	0.27
Fuller Park	63	60	25	0.92	0.72	0.60	4.20	3.75	1.67
Gage Park	14	46	104	0.93	0.72	0.29	0.93	2.88	6.93
Garfield Ridge	26	36	39	0.65	0.72	0.56	1.73	2.25	2.60
Grand Boulevard	684*	652*	141	0.91	0.68	0.54	45.60	40.75	9.40
Greater Grand Xing	222	296	268	0.82	0.65	0.46	14.80	18.50	17.87
Hegewisch	13	8	14	0.62	0.63	0.50	0.87	0.50	0.93
Hermosa	16	51	34	0.69	0.69	0.47	1.07	3.19	2.27
Humboldt Park	167	562*	341*	0.89	0.63	0.52	11.13	35.13	22.73
Hyde Park	62	72	18	0.69	0.61	0.67	4.13	4.50	1.20
Irving Park	28	61	42	0.71	0.67	0.52	1.87	3.81	2.80
Jefferson Park	4	10	12	0.50	0.40	1.00	0.27	0.63	0.80
Kenwood	181	125	25	0.92	0.59	0.44	12.07	7.81	1.67
Lake View	178	220	34	0.90	0.62	0.79	11.87	13.75	2.27
Lincoln Park	157	101	16	0.80	0.54	0.69	10.47	6.31	1.07
Lincoln Square	33	61	14	0.70	0.62	0.64	2.20	3.81	0.93
Logan Square	150	348	119	0.85	0.70	0.66	10.00	21.75	7.93
Loop	70	68	13	0.86	0.66	0.69	4.67	4.25	0.87
Lower West Side	180	283	100	0.85	0.58	0.52	12.00	17.69	6.67
Mckinley Park	16	23	21	1.00	0.70	0.48	1.07	1.44	1.40
Montclare	3	6	7	0.67	0.67	0.57	0.20	0.38	0.47
Morgan Park	61	79	61	0.89	0.67	0.49	4.07	4.94	4.07
Mount Greenwood	7	7	2	0.71	0.57	1.00	0.47	0.44	0.13
Near North Side	348	279	85	0.81	0.67	0.79	23.20	17.44	5.67
Near South Side	66	97	31	0.92	0.73	0.84	4.40	6.06	2.07
Near West Side	750*	695*	156	0.93	0.65	0.62	50.00	43.44	10.40
New City	121	390	252	0.89	0.72	0.50	8.07	24.38	16.80
North Center	23	52	16	1.00	0.71	0.88	1.53	3.25	1.07
North Lawndale	679*	577*	294	0.94	0.66	0.56	45.27	36.06	19.60
North Park	5	18	7	0.60	0.72	0.29	0.33	1.13	0.47
Norwood Park	9	13	7	0.78	0.69	1.00	0.60	0.81	0.47
Oakland	140	118	15	0.97	0.70	0.67	9.33	7.38	1.00
Ohare	14	13	3	0.36	0.77	0.67	0.93	0.81	0.20
Portage Park	16	25	25	0.69	0.56	0.64	1.07	1.56	1.67
Pullman	17	53	24	0.76	0.72	0.46	1.13	3.31	1.60
Riverdale	71	119	51	0.87	0.68	0.41	4.73	7.44	3.40
Rogers Park	43	179	72	0.84	0.60	0.54	2.87	11.19	4.80
Roseland	196	337	255	0.87	0.68	0.48	13.07	21.06	17.00
South Chicago	115	254	203	0.88	0.65	0.52	7.67	15.88	13.53
South Deering	39	72	63	0.82	0.69	0.57	2.60	4.50	4.20
South Lawndale	177	365	234	0.84	0.59	0.43	11.80	22.81	15.60
South Shore	323	473	297	0.82	0.60	0.50	21.53	29.56	19.80
Uptown	302	359	73	0.90	0.69	0.55	20.13	22.44	4.87
Washington Heights	82	116	114	0.84	0.67	0.39	5.47	7.25	7.60
Washington Park	386	364	111	0.86	0.60	0.45	25.73	22.75	7.40
West Elsdon	3	9	15	1.00	0.78	0.60	0.20	0.56	1.00
West Englewood	244	488	336*	0.93	0.73	0.43	16.27	30.50	22.40
West Garfield Park	313	350	219	0.84	0.63	0.50	20.87	21.88	14.60
West Lawn	12	7	27	0.58	0.43	0.67	0.80	0.44	1.80
West Pullman	67	227	193	0.99	0.65	0.46	4.47	14.19	12.87
West Ridge	11	34	35	0.82	0.71	0.60	0.73	2.13	2.33
West Town	423	545	136	0.84	0.59	0.54	28.20	34.06	9.07
Woodlawn	468	315	139	0.88	0.68	0.51	31.20	19.69	9.27
City totals and means	10,238	13,522	7,385	0.83	0.65	0.57	682.53	845.13	492.3

Sources. ¹Block et al. (2005); ²Chicago Open Data Portal Data.

Notes an outlier z-score, ≤2 SD

The results of the second analysis of variance (not shown) reveal that for homicides that occurred between 2001 and 2015, time to arrest was not a significant factor as there was no significant difference in clearance by arrest at 4 months (0.50), 32 months (0.51), or 39 months after the homicide (0.52; F = 0.194, p > .05, R² = 0.01, p = .08). Thus, time to arrest cannot explain the significantly lower arrest clearance percentage in the 2001 to 2015 time period as compared with 1965 to 1979 and 1980 to 1995. As seen in Table 1, this trend has continued precipitously lower over the past several years. Just 32% of homicides that occurred in 2016 and 25% of homicides that occurred in 2017 had been cleared by arrest as of March, 2019. Overall, just 28% of homicides that occurred between January 1, 2016 and August 2018 had been cleared through arrest by March, 2019, further limiting the deterrent effect of arrest on homicides in Chicago.

Regression Analysis

As homicide clearance by arrest continues to decline in Chicago, it is important to examine the community level predictors to see if the past research still holds true in Chicago or if predictors have also changed. To determine this, a series of regression analyses were conducted (see Table 3 below). While all six models are presented sequentially, the results of the sixth model are expanded upon here for clarity. In this model, the standardized arrest percentage for non-murder gun crime between 2007 and 2015, another measure of justice, was not a significant predictor of homicide clearance, controlling for other factors. Overall, this final model (Model 6, Table 3) explains just 52% of the variance across Community Areas in homicide arrest percentage over 15 years. Only two variables—the percentage of a Community Area population that is black, and the gang index—remain significant. The former has the strongest standardized coefficient (β = −0.79) and is negative, meaning that Community Areas with the highest black percentage corresponded with the lowest clearance rates by arrest, controlling for their gang/violent environment, economic hardship, gun violence clearance, life expectancy, and percent Hispanic population. The gang index was also significant and negative, suggesting that as gang and offender violence measures increase in a Community Area, the homicide arrest clearance rate decreases. However, this relationship is weak and close to being susceptible to a Type I error (saying the relationship is significant when it is not). The most important finding may be what is missing, that is, the at least 48% of the variance in homicide clearance by arrest that is not explained by the factors utilized.

Table 3.

OLS Regression Results Explaining Homicide Clearance through Arrest, 2001 to 2015, by Community Area.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
	β (SE)	β (SE)	β (SE)	β (SE)	β (SE)	β (SE)
Hardship index	−0.57*** (0.001)	−0.17 (0.001)	−0.10 (0.001)	0.04 (0.001)	−0.06 (0.001)	−0.07 (0.001)
Z avg gun crime rate 08–15	−0.09 (0.02)	0.05 (0.02)	0.14 (0.02)	0.23 (0.02)	0.18 (0.02)	0.17 (0.02)
Percent Black		−0.71*** (0.08)	−0.86*** (0.08)	−0.74*** (0.09)	−0.80*** (0.09)	−0.79*** (0.09)
Percent Hispanic		−0.55*** (0.09)	−0.29 (0.13)	−0.23 (0.13)	−0.18 (0.13)	−0.19 (0.13)
% English less than well			−0.35* (0.19)	−0.39* (0.19)	−0.31 (0.19)	−0.31 (0.19)
% English less than well and Spanish speaking			−0.12 (0.08)	−0.15 (0.08)	−0.18 (0.08)	−0.18 (0.08)
Gang, violence, drug offender index				−0.31 (0.09)	−0.44* (0.09)	−0.42* (0.10)
2010 life expectancy (fatalism)					−0.38* (0.01)	−0.37 (0.01)
Z gun crime arrest % 07–15						0.04 (0.02)
F	22.92***	17.35***	12.96***	11.72***	11.23***	9.87***
R² (change)	0.37	0.47 (0.11)	0.49 (0.02)	0.50 (0.01)	0.52 (0.02)	0.52 (0.00)
Durbin–Watson						2.4

p < .05. ***p < .001.

One limitation of the regression analysis sheds additional light on the findings. While all assumptions were met for the model, multicollinearity (high correlation between independent variables) provides further evidence of the co-occurrence of many factors within the black population of Chicago. While no models saw variance inflation factors greater than 10, some did exceed 5, meriting further examination. The percentage of black population in a Community Area showed high and positive correlations with the gang/violence environment index (r = 0.77), the z-average gun crime rate from 2008 through 2015 (r = 0.75) and the hardship index (r = 0.46). Furthermore, the black population percentage had high and negative correlations with the life expectancy (r = −0.83), percentage Hispanic (r = −0.59), and the percentage that speak English less than well (r = −0.63). While only one variable exceeds the threshold of 0.80 indicative of strong potential multicollinearity, suggesting the models are robust, these relationships point to co-occurring factors in Chicago’s black population that shed light on the lower murder arrest clearance in black Community Areas.

Limitations

There are several limitations of the study that are important to note. The first relates to the unit of analysis, which is a result of several other limitations. Previous research has used one extensive, yet now outdated, dataset that included individual level variables on the victims and/or offenders, but this data source does not. Thus, previous studies used an individual level of analysis or multilevel methods, which this dataset did not permit. The community area level in Chicago has been used in previous research and the issues in Chicago are widespread relative to CPD and its approach to homicide (see US Department of Justice, 2017), mitigating this limitation and adding to the existing literature, which has only been studied once since the mid-2000’s in Chicago using a slightly different geography of neighborhood clusters (Mancik et al., 2018). It is likely that some proportion of the unexplained variance at the community level is found in individual factors, but no inferences are made based on the present results.

No measures of police involvement were included in this study as the CPD does not produce enough public information about the assignment of detectives to cases, the total number of detectives or other manpower measures related to homicide investigation to include this as a variable. It is likely that a certain percentage of the unexplained variance is due to CPD factors, but this was not addressed in the current study. To mitigate this limitation, no inferences are made about police devaluation or factors beyond the community itself.

The time frames utilized from the end of 2015 (4, 30, and 39 months) were not statistically derived based on the prior literature, but rather when follow-up was permitted. From the second analysis of variance, it seems as if a scheduled follow-up would likely not change the results and the follow-up period extended almost 4 years after the last homicide and 18 years after the first-occurring in the sample. Future research seeking to identify specific cut-points of clearance outcomes may use a more determinative schedule for specificity.

Discussion and Conclusion

This research used a new data source, an updated time frame and adds to the literature on homicide clearance by arrest in Chicago as well as using the community area as the unit of analysis. The results of the present study show that homicide clearance by arrest in Chicago was significantly lower in the most recent time period (2001–2015) when compared to 1965 to 1979 and 1980 to 1995. Most of the previous studies in Chicago studied earlier time periods. The trend in recent years has continued downward to much lower levels. All graphic evidence shows this continual decline in justice for homicide victims over time in all aspects. As an example, for homicides that occurred in 2009, offenders had the same odds as flipping a coin as to whether they would be arrested for the homicides they committed even up to after 10 years after the homicide was committed. Even the advent of time for recent time period homicides at 4-, 30-, and 39-month intervals does not show significant increases of homicide clearance by arrest, with 1.9% of homicides additionally cleared by arrest over 3 years later for a 15-year period.

Though not studied directly, the significant decline in homicide arrest rates occurred during a time of significant technological expansion in cameras and linked ShotSpotter sensors that came at great cost but seem to have little statistical impact on arrests for homicides. Ten years before homicide reached an epidemic level in Chicago in 2016 (see Ferrandino, 2018), Rosenbaum and Stevens (2005) reported that the CPD viewed increased use of surveillance cameras as an effective deterrent strategy for illegal activity in Chicago. One such strategy was “Operation Disruption,” which used marked and visible cameras to deter violence, with CPD claiming a 17% drop in violent crime in areas the cameras were deployed. Thirty units were operational in 2004 and 50 more were anticipated by 2005. The CPD reported 928 in operation by 2010 (Chicago Police Department, 2010). There are now over 30,000 cameras strategically placed in the city (Williams, 2018), up from 10,000 in 2011, that cost the city over $60 million; 1,260 of these cameras were Police Observation Devices, or POD’s, owned by the Chicago Police Department in addition to thousands of cameras in schools, transit locations, and private property (American Civil Liberties Union, 2011). The camera network is integrated with a ShotSpotter gunfire acoustic detection system that will soon cover 100 square miles and cost $23 million over 3 years (ShotSpotter, 2018). In theory, these technological tools should lead to more homicides being cleared without the need for witnesses or other evidence, which may have contributed to the decline in homicide clearance rates (see Jiao, 2007; Litwin & Xu, 2007). Cameras are important as murder has become much more common in Chicago outdoors (84% of murders in 2011 compared with 61% in 1991) and in public spaces (68% in 2011 compared with 48% in 1991). Shotspotter has the ability to play an important role as 83% of murder victims in 2011 were shot, increasing to 92% by 2017 (see Chicago Police Department, 2011, 2017). Despite this great investment in technology, clearance rates by arrest are trending downward in Chicago, outpacing the same national trend. Future research on the role of technology in clearing homicides in Chicago is essential given the financial and social costs (American Civil Liberties Union, 2011) amid declining clearance rates.

The regression models are important for several reasons. First, contrary to previous research which found Hispanics to be a significant negative predictor of homicide clearance at both the victim (Alderden & Lavery, 2007; Litwin, 2004; Litwin & Xu, 2007; Xu, 2008) and community area levels (Litwin & Xu, 2007), the final regression model shows the risk factor most related to arrest clearance now is the percentage of a Community Area’s population which is black. This finding at the Community Area level also differs from previous research that found the percentage of black population was not a factor in homicide clearance (Litwin & Xu, 2007) and that the highest clearance rates by arrest in Chicago over the same time period are found in the highest homicide census block groups where Blacks comprise over 90% of the population (Ferrandino, 2018). Some of this difference could be attributed to the previous research largely being conducted at multiple levels or the individual level but not the community level, as well as the utilization of a different data source and time frame. However, the findings here suggest that arrest clearance decrease for murder is far more acute and potentially detrimental in Community Areas with high proportion of black population, compared to those with a high proportion of Latino population.

The US Department of Justice (2017) investigation into the CPD found that the decrease in homicide clearance was attributable to the CPD’s relationship with the minority community generally. The findings here suggest it is far more acute, and thus far more potentially detrimental, in Chicago’s black communities than in its Hispanic communities. This is an important distinction, as the US Department of Justice (2017) states homicide clearance is “one of the most critical problems currently facing Chicago” (p. 153), yet family members of homicide victims reported “their experience with CPD, after a family member had been murdered, had made them feel that CPD does not genuinely care about the murders of young black men and women, and do too little to investigate and resolve those homicides” (p. 143). The present study finds the statistical relationship between Chicago’s black population and lower clearance rates to be true. However, the study cannot shed light on the reasons for this as police devaluation or other potential causes were not specifically conceptualized, measured, or assessed.

The limitations of the regression model shed further light on the main findings related to race and murder arrest clearance in Chicago. The Community Areas with higher percentages of black population are racially isolated (r = −0.60 between black and Hispanic population percentages), face stronger economic hardships than Hispanic communities, face a harsher relative gun violence rate, are more impacted by the gang/violence environment and face a strong, negative relationship with life expectancy (r = −0.83) that is opposite of the Hispanic community (r = 0.39). Taken together, these co-occurring problems in the Black population of Chicago, at the Community Area level, show the interaction of these social factors in ways that may have manifested as a lower arrest clearance rate for murder in the Windy City, explaining roughly half of the variance in these percentages across Chicago. Future research is needed to determine how, and it what ways, these symptoms interact and contribute to the statistically lower homicide clearance rates found in Chicago in the Community Areas with larger black populations that differs from the other populations in the city. Such results would be critical in determining what actions can be done to alleviate this problem and start making more arrests for murder in Chicago.

Perhaps the most important finding is what was missing. Just over half of the variance in homicide clearance by arrest in Chicago was explained by the two variables. As only two measures are significant in the final model, the r-square value is likely inflated, meaning there are several other important but unmeasured factors that are critical for future research to explore, including, for example, detective assignment, workload, technology utilization, and other city policies. Many factors expected to be related to homicide clearance turned out not to be statistically significant, including: hardship index, average gun crime rate, average gun crime arrest rate, the percent Hispanic, the percentage that speak English less than well, the percentage that speak English less than well and speak Spanish and the life expectancy. Homicide clearance by arrest is plunging in Chicago, especially in black communities, even when the gang/violent environment has been taken into account. Addressing this salient issue would enhance justice in the city by holding offenders accountable and perhaps begin to improve views toward the police in Chicago’s black communities.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Joseph Ferrandino

Notes

Author Biography

Joseph Ferrandino is the director of Criminal Justice at Ferris State University in Big Rapids, MI. His research interests focus on stop and frisk in NYC, police policy analysis and teaching criminal justice. He has also consulted with criminal justice agencies from the local to federal levels.

References

Alderden

M. A.

Lavery

T. A.

(2007). Predicting homicide clearances in Chicago: Investigating disparities in predictors across different types of homicide. Homicide Studies, 11(2), 115–132.

American Civil Liberties Union. (2011). Chicago’s video surveillance cameras: A pervasive and unregulated threat to our privacy. ACLU. https://www.aclu-il.org/en/publications/video-camera-surveillance-chicago. (accessed 1 June 2018).

Baskin

Sommers

(2010). The influence of forensic evidence on the case outcomes of homicide incidents. Journal of Criminal Justice, 38(6), 1141–1149.

Block

C. R.

Block

R. L.

, & Illinois Criminal Justice Information Authority. (2005). Homicides in Chicago, 1965–1995. Inter-university Consortium for Political and Social Research.

Block

(1976). Homicide in Chicago: A nine-year study (1965–1973). Journal of Criminal Law and Criminology, 66(4), 496–510.

Borg

M. J.

Parker

K. F.

(2001). Mobilizing law in urban areas: The social structure of homicide clearance rates. Law & Society Review, 35(2), 435–466.

Chicago Data Portal. (2014a). Census data—Languages spoken in Chicago, 2008—2012. https://data.cityofchicago.org/Health-Human-Services/Census-Data-Languages-spoken-in-Chicago-2008-2012/a2fk-ec6q. (accessed 5 April 2019).

Chicago Data Portal. (2014b). Census data—Selected socioeconomic indicators in Chicago, 2008–2012. https://data.cityofchicago.org/Health-Human-Services/Census-Data-Selected-socioeconomic-indicators-in-C/kn9c-c2s2. (accessed 4 April 2019).

Chicago Data Portal. (2014c). Public health statistics—Life expectancy by community area. https://data.cityofchicago.org/Health-Human-Services/Public-Health-Statistics-Life-Expectancy-By-Commun/qjr3-bm53. (accessed 5 April 2019).

10.

Chicago Data Portal. (2017). Strategic subject list. https://data.cityofchicago.org/Public-Safety/Strategic-Subject-List/4aki-r3np. (accessed 5 April 2019).

11.

Chicago Data Portal. (2020). Crimes—2001 to present. https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-present/ijzp-q8t2

12.

Chicago Police Department. (2011). Chicago murder analysis. https://home.chicagopolice.org/2011-murder-report/. (accessed 11 April 2016).

13.

Chicago Police Department. (2010). Annual report 2010: A year in review. https://home.chicagopolice.org/wp-content/uploads/2014/12/2010-Annual-Report.pdf. (accessed 1 July 2017).

14.

Chicago Police Department. (2017). Annual report. https://home.chicagopolice.org/statistics-data/statistical-reports/annual-reports/. (accessed 24 September 2018).

15.

Encyclopedia of Chicago. (n.d.). Chicago’s community areas. http://www.encyclopedia.chicagohistory.org/pages/1760.html

16.

FBI. (n.d.). UCR data tool. https://www.ucrdatatool.gov/Search/Crime/Crime.cfm

17.

FBI Uniform Crime Reports. (2013, 2014). Crime in the United States. https://www.fbi.gov/services/cjis/ucr/publications

18.

Ferrandino

(2018). Endemic, outbreak or epidemic? Geographies of affliction, exposure and homicide immunity in Chicago. Journal of Crime and Justice, 41(4), 347–363.

19.

Illinois Uniform Crime Reports (2017). Monthly index crime form. http://www.isp.state.il.us/crime/ucrhome.cfm

20.

Jiao

A. Y.

(2007). Explaining homicide clearance: An analysis of Chicago homicide data 1965–1995. Criminal Justice Studies, 20(1), 3–14.

21.

Litwin

K. J.

(2004). A multilevel multivariate analysis of factors affecting homicide clearances. Journal of Research in Crime and Delinquency, 41(4), 327–351.

22.

Litwin

K. J.

(2007). The dynamic nature of homicide clearances: A multilevel model comparison of three time periods. Homicide Studies, 11(2), 94–114.

23.

Mancik

A. M.

Parker

K. F.

Williams

K. R.

(2018). Neighborhood context and homicide clearance: Estimating the effects of collective efficacy. Homicide Studies, 22(2), 188–213.

24.

Ousey

G. C.

Lee

M. R.

(2010). To know the unknown: The decline in homicide clearance rates, 1980–2000. Criminal Justice Review, 35(2), 141–158.

25.

Peterson

(2015). Neighbourhood context and unsolved murders: The social ecology of homicide investigations. Policing and Society: An International Journal of Research and Policy, 27(4), 372–392.

26.

Puckett

J. L.

Lundman

R. J.

(2003). Factors affecting homicide clearances: Multivariate analysis of a more complete conceptual framework. Journal of Research in Crime and Delinquency, 40(2), 171–193.

27.

Regoeczi

W. C.

Jarvis

J. P.

(2013). Beyond the social production of homicide rates: Extending social disorganization theory to explain homicide case outcomes. Justice Quarterly, 30(6), 983–1014.

28.

Regoeczi

W. C.

Jarvis

Riedel

(2008). Clearing murders: Is it about time? Journal of Research in Crime & Delinquency, 45(2), 142–162.

29.

Regoeczi

W. C.

Kennedy

L. L.

Silverman

R. A.

(2000). Uncleared homicides: A Canada/United States Comparison. Homicide Studies, 4(2), 135–161.

30.

Riedel

(2008). Homicide arrest clearances: A review of the literature. Sociology Compass, 2(4), 1145–1164.

31.

Riedel

Boulahanis

J. G.

(2007). Homicides exceptionally cleared and cleared by arrest: An exploratory study of police/prosecutor outcomes. Homicide Studies, 11(2), 151–164.

32.

Riedel

Rinehart

T. A.

(1996). Murder clearances and missing data. Journal of Crime and Justice, 19(2), 83–102.

33.

Riedel

Jarvis

(1999). The decline of arrest clearances for criminal homicides: Causes, correlates and third parties. Criminal Justice Policy Review, 9(3/4); 279–306.

34.

Roberts

Lyons

C. J.

(2011). Hispanic victims and homicide clearance by arrest. Homicide Studies, 15(1), 48–73.

35.

Rosenbaum

D. P.

Stevens

(2005). In University of Illinois at Chicago Center for Research in Law and Justice (Eds.) Reducing public violence and homicide in Chicago: Strategies and tactics of the Chicago Police Department. Illinois Criminal Justice Information Authority.

36.

Rydberg

Pizarro

J. M.

(2014). Victim lifestyle as a correlate of homicide clearance. Homicide Studies, 18(4), 342–362.

37.

ShotSpotter. (2018). Chicago signs $23 million multi-year agreement with Shotspotter to extend gunshot detection coverage into next decade. ShotSpotter, Inc. https://globenewswire.com/news-release/2018/09/05/1565583/0/en/Chicago-Signs-23-Million-Multi-year-Agreement-With-Shotspotter-to-Extend-Gunshot-Detection-Coverage-Into-Next-Decade.html. (accessed 24 April 2019).

38.

US Department of Justice. (2017). Investigation of the Chicago police department. Civil Rights Division and United States Attorney’s Office Northern District of Illinois.

39.

Williams

(2018, May 26). Can 30,000 cameras help solve Chicago’s crime problem? Armed with advanced gadgets and mapping, officers can get to crime scenes “in time to see the guy still shooting.” But what does it mean for residents’ privacy? New York Times. https://www.nytimes.com/2018/05/26/us/chicago-police-surveillance.html. (accessed 1 July 2018).

40.

(2008). The characteristics of homicide events and the decline in homicide clearance: A longitudinal approach to the dynamic relationship, Chicago 1966-1995. Criminal Justice Review, 33(4), 453–479.